summaryrefslogtreecommitdiff
path: root/src/network
AgeCommit message (Collapse)Author
2010-05-13(svn r19814) -Codechange: give some more unnamed enums a name, in case they ↵rubidium
consisted of unrelated values use static const (u)int
2010-05-13(svn r19809) -Codechange: make some unnamed network related enums/defines ↵rubidium
static const variables
2010-05-13(svn r19808) -Codechange: NetworkCalculateLag returned an uint, so keep it ↵rubidium
that way
2010-05-10(svn r19781) -Fix [FS#3809]: compilation on NetBSD failed (Krille)rubidium
2010-05-10(svn r19777) -Change: use the file scanner to find the .tarsrubidium
2010-04-24(svn r19714) -Feature: ctrl+click on a vehicle to start/stop itsmatz
2010-04-22(svn r19695) -Fix: leaking a file descriptorrubidium
2010-04-22(svn r19693) -Codechange: split STATUS_INACTIVE to two statessmatz
2010-04-21(svn r19686) -Fix (r15126): truncated archives were not detected when using ↵smatz
zlib 1.2.3. This also fixes zlib 1.2.4 compatibility, zlib 1.2.5 is bugfree
2010-04-20(svn r19680) -Fix [FS#3775] (r19648): when joining a MP game all clients ↵rubidium
with company ID > 0 would be shown as if they were a spectator
2010-04-19(svn r19678) -Fix (r19607): client status was shown incorrect in the consolerubidium
2010-04-17(svn r19651) -Fix [FS#3745]: when a company is sold, move connected clients ↵smatz
to spectators
2010-04-17(svn r19648) -Fix [FS#3760]: a client would not be properly moved when moved ↵rubidium
while joining, e.g. when entering a company's password. This caused the client to be in the wrong company (according to the rest of the clients) and the client being kicked on the first command
2010-04-14(svn r19629) -Fix [FS#3757]: chat message caused glitch when rejoining a ↵rubidium
network game
2010-04-13(svn r19620) -Fix: desync when a command is received and in the queue while ↵rubidium
a client starts joining, i.e. save the game state. This can happen in two ways: with frame_freq > 1 a command received in a previous frame might not be executed yet or when a command is received in the same frame as the join but before the savegame is made. In both cases the joining client would not get all commands to get in-sync with the server (and the other clients).
2010-04-13(svn r19619) -Fix (r19618): [desync debug] inserting the "join" pause could ↵rubidium
cause a crash as some command data was not properly initialised
2010-04-13(svn r19617) -Fix [desync debug]: log the sync state only once per day, not ↵rubidium
multiple times when paused with _date_fract = 0 -Change [desync debug]: check the sync state from the command stream and make sure no unknown input is encountered
2010-04-12(svn r19614) -Codechange: "it's" => "its" where appropriatesmatz
2010-04-11(svn r19613) -Fix [FS#3755]: possible invalid read when server moves client ↵smatz
to spectators before he finishes joining
2010-04-11(svn r19610) -Codechange: rename STATUS_AUTH to STATUS_AUTHORIZEDsmatz
2010-04-11(svn r19609) -Fix: do not print 'Connection lost' message for clients who ↵smatz
failed to enter correct password
2010-04-11(svn r19608) -Fix: do not kick client if he entered password and the ↵smatz
password was cleared meanwhile
2010-04-11(svn r19607) -Codechange: use different packet types instead of packet subtypessmatz
2010-04-11(svn r19606) -Codechange: remove global variable pw_typesmatz
2010-04-11(svn r19602) -Codechange: add the name of the command to the desync logsrubidium
2010-04-11(svn r19601) -Fix: make the desync replay parse the "whole" string instead ↵rubidium
of everything up to the first space; make it log the sync state too
2010-04-10(svn r19600) -Fix (r18994): Presence of online content was not properly ↵frosch
updated after download due to duplicate slashes in the path.
2010-04-10(svn r19596) -Fix (r19593): dedicated server always fast forwarding if ↵glx
compiled without DEBUG_DUMP_COMMANDS
2010-04-09(svn r19593) -Codechange: improve loading/executing the desync logsrubidium
2010-04-08(svn r19589) -Change: add some more useful information to the desync log and ↵rubidium
unify the formatting
2010-03-23(svn r19508) -Codechange: remove semicolon after ↵smatz
DEFINE_UNAVAILABLE_CONTENT_RECEIVE_COMMAND and DEFINE_UNAVAILABLE_UDP_RECEIVE_COMMAND
2010-03-23(svn r19507) -Codechange: remove semicolon after DECLARE_POSTFIX_INCREMENT ↵smatz
and DECLARE_ENUM_AS_BIT_SET
2010-03-14(svn r19423) -Codechange: Move error message reporting of industry build in ↵alberth
SE to a callback.
2010-03-11(svn r19385) -Fix [FS#3684]: [IPv6] Netmask calculations were wrong if cidr ↵rubidium
>= 32.
2010-03-08(svn r19374) -Change: Update documentation for console command connect to ↵peter1138
use ip:port#company parameter format, in line with command line help.
2010-02-28(svn r19295) -Codechange: introduce wrapper functions for GRFConfig::name/infoyexo
2010-02-27(svn r19283) -Fix (r19256): GRFConfig is not zeroed for non-dynamic ↵frosch
construction. GRFIdentifier is sufficient though.
2010-02-26(svn r19262) -Feature: finalise BaNaNaS support for music setsrubidium
2010-02-25(svn r19256) -Codechange: use a constructor/destructor for GRFConfig to make ↵yexo
sure all members are properly initialized
2010-02-25(svn r19255) -Codechange: encapsulate GRFIdentifier in GRFConfig instead of ↵yexo
subclassing it
2010-02-24(svn r19227) -Codechange: Reorganization of parameters at CmdBuildRoadStop.terkhen
2010-02-24(svn r19224) -Codechange: change parameters passed to ShowErrorMessage() a bitsmatz
2010-02-14(svn r19132) -Codechange: No need to end a line with ;;.alberth
2010-02-11(svn r19094) -Codechange: don't check for NULL values before calling StrEmptyyexo
2010-02-10(svn r19090) -Codechange: some comment coding style fixesrubidium
2010-02-10(svn r19087) -Codechange: some typos in StringIDsrubidium
2010-02-10(svn r19081) -Codechange: make it possible to disable compilation of the ↵rubidium
AI+Squirrel
2010-02-10(svn r19075) -Codechange: unhackify NetworkChangeCompanyPassword()smatz
2010-02-09(svn r19072) -Fix [FS#3599]: possible read/write after free when the client ↵rubidium
triggered the server to close the connection
2010-02-07(svn r19054) -Fix [FS#3598]: when banning yourself via rcon do not send the ↵rubidium
'command response' to the client as the connection has already been terminated